Structure-independent machine-learning predictions of the CDK12 interactome.

Cyclin-dependent kinase 12 (CDK12) is a critical regulatory protein involved in transcription and DNA repair processes. Dysregulation of CDK12 has been implicated in various diseases, including cancer. Understanding the CDK12 interactome is pivotal for elucidating its functional roles and potential therapeutic targets. Traditional methods for interactome prediction often rely on protein structure information, limiting applicability to CDK12 characterized by partly disordered terminal C region. In this study, we present a structure-independent machine-learning model that utilizes proteins' sequence and functional data to predict the CDK12 interactome. This approach is motivated by the disordered character of the CDK12 C-terminal region mitigating a structure-driven search for binding partners. Our approach incorporates multiple data sources, including protein-protein interaction networks, functional annotations, and sequence-based features, to construct a comprehensive CDK12 interactome prediction model. The ability to predict CDK12 interactions without relying on structural information is a significant advancement, as many potential interaction partners may lack crystallographic data. In conclusion, our structure-independent machine-learning model presents a powerful tool for predicting the CDK12 interactome and holds promise in advancing our understanding of CDK12 biology, identifying potential therapeutic targets, and facilitating precision-medicine approaches for CDK12-associated diseases.

Bayesian network models identify cooperative GPCR:G protein interactions that contribute to G protein coupling.

Cooperative interactions in protein-protein interfaces demonstrate the interdependency or the linked network-like behavior and their effect on the coupling of proteins. Cooperative interactions also could cause ripple or allosteric effects at a distance in protein-protein interfaces. Although they are critically important in protein-protein interfaces, it is challenging to determine which amino acid pair interactions are cooperative. In this work, we have used Bayesian network modeling, an interpretable machine learning method, combined with molecular dynamics trajectories to identify the residue pairs that show high cooperativity and their allosteric effect in the interface of G protein-coupled receptor (GPCR) complexes with Gα subunits. Our results reveal six GPCR:Gα contacts that are common to the different Gα subtypes and show strong cooperativity in the formation of interface. Both the C terminus helix5 and the core of the G protein are codependent entities and play an important role in GPCR coupling. We show that a promiscuous GPCR coupling to different Gα subtypes, makes all the GPCR:Gα contacts that are specific to each Gα subtype (Gαs, Gαi, and Gαq). This work underscores the potential of data-driven Bayesian network modeling in elucidating the intricate dependencies and selectivity determinants in GPCR:G protein complexes, offering valuable insights into the dynamic nature of these essential cellular signaling components.

Regulation of Enhancers by SUMOylation Through TFAP2C Binding and Recruitment of HDAC Complex to the Chromatin.

Enhancers are fundamental to gene regulation. Post-translational modifications by the small ubiquitin-like modifiers (SUMO) modify chromatin regulation enzymes, including histone acetylases and deacetylases. However, it remains unclear whether SUMOylation regulates enhancer marks, acetylation at the 27th lysine residue of the histone H3 protein (H3K27Ac). To investigate whether SUMOylation regulates H3K27Ac, we performed genome-wide ChIP-seq analyses and discovered that knockdown (KD) of the SUMO activating enzyme catalytic subunit UBA2 reduced H3K27Ac at most enhancers. Bioinformatic analysis revealed that TFAP2C-binding sites are enriched in enhancers whose H3K27Ac was reduced by UBA2 KD. ChIP-seq analysis in combination with molecular biological methods showed that TFAP2C binding to enhancers increased upon UBA2 KD or inhibition of SUMOylation by a small molecule SUMOylation inhibitor. However, this is not due to the SUMOylation of TFAP2C itself. Proteomics analysis of TFAP2C interactome on the chromatin identified histone deacetylation (HDAC) and RNA splicing machineries that contain many SUMOylation targets. TFAP2C KD reduced HDAC1 binding to chromatin and increased H3K27Ac marks at enhancer regions, suggesting that TFAP2C is important in recruiting HDAC machinery. Taken together, our findings provide insights into the regulation of enhancer marks by SUMOylation and TFAP2C and suggest that SUMOylation of proteins in the HDAC machinery regulates their recruitments to enhancers.

Transcriptome free energy can serve as a dynamic patient-specific biomarker in acute myeloid leukemia.

Acute myeloid leukemia (AML) is prevalent in both adult and pediatric patients. Despite advances in patient categorization, the heterogeneity of AML remains a challenge. Recent studies have explored the use of gene expression data to enhance AML diagnosis and prognosis, however, alternative approaches rooted in physics and chemistry may provide another level of insight into AML transformation. Utilizing publicly available databases, we analyze 884 human and mouse blood and bone marrow samples. We employ a personalized medicine strategy, combining state-transition theory and surprisal analysis, to assess the RNA transcriptome of individual patients. The transcriptome is transformed into physical parameters that represent each sample's steady state and the free energy change (FEC) from that steady state, which is the state with the lowest free energy.We found the transcriptome steady state was invariant across normal and AML samples. FEC, representing active molecular processes, varied significantly between samples and was used to create patient-specific barcodes to characterize the biology of the disease. We discovered that AML samples that were in a transition state had the highest FEC. This disease state may be characterized as the most unstable and hence the most therapeutically targetable since a change in free energy is a thermodynamic requirement for disease progression. We also found that distinct sets of ongoing processes may be at the root of otherwise similar clinical phenotypes, implying that our integrated analysis of transcriptome profiles may facilitate a personalized medicine approach to cure AML and restore a steady state in each patient.

State-transition modeling of blood transcriptome predicts disease evolution and treatment response in chronic myeloid leukemia.

Chronic myeloid leukemia (CML) is initiated and maintained by BCR::ABL which is clinically targeted using tyrosine kinase inhibitors (TKIs). TKIs can induce long-term remission but are also not curative. Thus, CML is an ideal system to test our hypothesis that transcriptome-based state-transition models accurately predict cancer evolution and treatment response. We collected time-sequential blood samples from tetracycline-off (Tet-Off) BCR::ABL-inducible transgenic mice and wild-type controls. From the transcriptome, we constructed a CML state-space and a three-well leukemogenic potential landscape. The potential's stable critical points defined observable disease states. Early states were characterized by anti-CML genes opposing leukemia; late states were characterized by pro-CML genes. Genes with expression patterns shaped similarly to the potential landscape were identified as drivers of disease transition. Re-introduction of tetracycline to silence the BCR::ABL gene returned diseased mice transcriptomes to a near healthy state, without reaching it, suggesting parts of the transition are irreversible. TKI only reverted the transcriptome to an intermediate disease state, without approaching a state of health; disease relapse occurred soon after treatment. Using only the earliest time-point as initial conditions, our state-transition models accurately predicted both disease progression and treatment response, supporting this as a potentially valuable approach to time clinical intervention, before phenotypic changes become detectable.

Systems profiling reveals recurrently dysregulated cytokine signaling responses in ER+ breast cancer patients' blood.

Cytokines mediate cell-to-cell communication across the immune system and therefore are critical to immunosurveillance in cancer and other diseases. Several cytokines show dysregulated abundance or signaling responses in breast cancer, associated with the disease and differences in survival and progression. Cytokines operate in a coordinated manner to affect immune surveillance and regulate one another, necessitating a systems approach for a complete picture of this dysregulation. Here, we profiled cytokine signaling responses of peripheral immune cells from breast cancer patients as compared to healthy controls in a multidimensional manner across ligands, cell populations, and responsive pathways. We find alterations in cytokine responsiveness across pathways and cell types that are best defined by integrated signatures across dimensions. Alterations in the abundance of a cytokine's cognate receptor do not explain differences in responsiveness. Rather, alterations in baseline signaling and receptor abundance suggesting immune cell reprogramming are associated with altered responses. These integrated features suggest a global reprogramming of immune cell communication in breast cancer.

Bayesian network models identify co-operative GPCR:G protein interactions that contribute to G protein coupling.

Cooperative interactions in protein-protein interfaces demonstrate the interdependency or the linked network-like behavior of interface interactions and their effect on the coupling of proteins. Cooperative interactions also could cause ripple or allosteric effects at a distance in protein-protein interfaces. Although they are critically important in protein-protein interfaces it is challenging to determine which amino acid pair interactions are cooperative. In this work we have used Bayesian network modeling, an interpretable machine learning method, combined with molecular dynamics trajectories to identify the residue pairs that show high cooperativity and their allosteric effect in the interface of G protein-coupled receptor (GPCR) complexes with G proteins. Our results reveal a strong co-dependency in the formation of interface GPCR:G protein contacts. This observation indicates that cooperativity of GPCR:G protein interactions is necessary for the coupling and selectivity of G proteins and is thus critical for receptor function. We have identified subnetworks containing polar and hydrophobic interactions that are common among multiple GPCRs coupling to different G protein subtypes (Gs, Gi and Gq). These common subnetworks along with G protein-specific subnetworks together confer selectivity to the G protein coupling. This work underscores the potential of data-driven Bayesian network modeling in elucidating the intricate dependencies and selectivity determinants in GPCR:G protein complexes, offering valuable insights into the dynamic nature of these essential cellular signaling components.

State-transition Modeling of Blood Transcriptome Predicts Disease Evolution and Treatment Response in Chronic Myeloid Leukemia.

Chronic myeloid leukemia (CML) is initiated and maintained by BCR::ABL which is clinically targeted using tyrosine kinase inhibitors (TKIs). TKIs can induce long-term remission but are also not curative. Thus, CML is an ideal system to test our hypothesis that transcriptome-based state-transition models accurately predict cancer evolution and treatment response. We collected time-sequential blood samples from tetracycline-off (Tet-Off) BCR::ABL-inducible transgenic mice and wild-type controls. From the transcriptome, we constructed a CML state-space and a three-well leukemogenic potential landscape. The potential's stable critical points defined observable disease states. Early states were characterized by anti-CML genes opposing leukemia; late states were characterized by pro-CML genes. Genes with expression patterns shaped similarly to the potential landscape were identified as drivers of disease transition. Re-introduction of tetracycline to silence the BCR::ABL gene returned diseased mice transcriptomes to a near healthy state, without reaching it, suggesting parts of the transition are irreversible. TKI only reverted the transcriptome to an intermediate disease state, without approaching a state of health; disease relapse occurred soon after treatment. Using only the earliest time-point as initial conditions, our state-transition models accurately predicted both disease progression and treatment response, supporting this as a potentially valuable approach to time clinical intervention even before phenotypic changes become detectable.

Acquired miR-142 deficit in leukemic stem cells suffices to drive chronic myeloid leukemia into blast crisis.

The mechanisms underlying the transformation of chronic myeloid leukemia (CML) from chronic phase (CP) to blast crisis (BC) are not fully elucidated. Here, we show lower levels of miR-142 in CD34+CD38- blasts from BC CML patients than in those from CP CML patients, suggesting that miR-142 deficit is implicated in BC evolution. Thus, we create miR-142 knockout CML (i.e., miR-142-/-BCR-ABL) mice, which develop BC and die sooner than miR-142 wt CML (i.e., miR-142+/+BCR-ABL) mice, which instead remain in CP CML. Leukemic stem cells (LSCs) from miR-142-/-BCR-ABL mice recapitulate the BC phenotype in congenic recipients, supporting LSC transformation by miR-142 deficit. State-transition and mutual information analyses of "bulk" and single cell RNA-seq data, metabolomic profiling and functional metabolic assays identify enhanced fatty acid ß-oxidation, oxidative phosphorylation and mitochondrial fusion in LSCs as key steps in miR-142-driven BC evolution. A synthetic CpG-miR-142 mimic oligodeoxynucleotide rescues the BC phenotype in miR-142-/-BCR-ABL mice and patient-derived xenografts.

Bow-tie architectures in biological and artificial neural networks: Implications for network evolution and assay design.

Modern artificial neural networks (ANNs) have long been designed on foundations of mathematics as opposed to their original foundations of biomimicry. However, the structure and function of these modern ANNs are often analogous to real-life biological networks. We propose that the ubiquitous information-theoretic principles underlying the development of ANNs are similar to the principles guiding the macro-evolution of biological networks and that insights gained from one field can be applied to the other. We generate hypotheses on the bow-tie network structure of the Janus kinase - signal transducers and activators of transcription (JAK-STAT) pathway, additionally informed by the evolutionary considerations, and carry out ANN simulation experiments to demonstrate that an increase in the network's input and output complexity does not necessarily require a more complex intermediate layer. This observation should guide novel biomarker discovery-namely, to prioritize sections of the biological networks in which information is most compressed as opposed to biomarkers representing the periphery of the network.

The Human Brainome: changes in expression of VGF, SPECC1L, HLA-DRA and RANBP3L act with APOE E4 to alter risk for late onset Alzheimer's disease.

While there are currently over 40 replicated genes with mapped risk alleles for Late Onset Alzheimer's disease (LOAD), the Apolipoprotein E locus E4 haplotype is still the biggest driver of risk, with odds ratios for neuropathologically confirmed E44 carriers exceeding 30 (95% confidence interval 16.59-58.75). We sought to address whether the APOE E4 haplotype modifies expression globally through networks of expression to increase LOAD risk. We have used the Human Brainome data to build expression networks comparing APOE E4 carriers to non-carriers using scalable mixed-datatypes Bayesian network (BN) modeling. We have found that VGF had the greatest explanatory weight. High expression of VGF is a protective signal, even on the background of APOE E4 alleles. LOAD risk signals, considering an APOE background, include high levels of SPECC1L, HLA-DRA and RANBP3L. Our findings nominate several new transcripts, taking a combined approach to network building including known LOAD risk loci.

Integration of single-cell transcriptomes and biological function reveals distinct behavioral patterns in bone marrow endothelium.

Heterogeneity of endothelial cell (EC) populations reflects their diverse functions in maintaining tissue's homeostasis. However, their phenotypic, molecular, and functional properties are not entirely mapped. We use the Tie2-CreERT2;Rosa26-tdTomato reporter mouse to trace, profile, and cultivate primary ECs from different organs. As paradigm platform, we use this strategy to study bone marrow endothelial cells (BMECs). Single-cell mRNA sequencing of primary BMECs reveals that their diversity and native molecular signatures is transitorily preserved in an ex vivo culture that conserves key cell-to-cell microenvironment interactions. Macrophages sustain BMEC cellular diversity and expansion and preserve sinusoidal-like BMECs ex vivo. Endomucin expression discriminates BMECs in populations exhibiting mutually exclusive properties and distinct sinusoidal/arterial and tip/stalk signatures. In contrast to arterial-like, sinusoidal-like BMECs are short-lived, form 2D-networks, contribute to in vivo angiogenesis, and support hematopoietic stem/progenitor cells in vitro. This platform can be extended to other organs' ECs to decode mechanistic information and explore therapeutics.

Dynamic patterns of microRNA expression during acute myeloid leukemia state-transition.

MicroRNAs (miRNAs) have been shown to hold prognostic value in acute myeloid leukemia (AML); however, the temporal dynamics of miRNA expression in AML are poorly understood. Using serial samples from a mouse model of AML to generate time-series miRNA sequencing data, we are the first to show that the miRNA transcriptome undergoes state-transition during AML initiation and progression. We modeled AML state-transition as a particle undergoing Brownian motion in a quasi-potential and validated the AML state-space and state-transition model to accurately predict time to AML in an independent cohort of mice. The critical points of the model provided a framework to align samples from mice that developed AML at different rates. Our mathematical approach allowed discovery of dynamic processes involved during AML development and, if translated to humans, has the potential to predict an individual's disease trajectory.

MicroRNA networks in FLT3-ITD acute myeloid leukemia.

MiR-126 and miR-155 are key microRNAs (miRNAs) that regulate, respectively, hematopoietic cell quiescence and proliferation. Herein we showed that in acute myeloid leukemia (AML), the biogenesis of these two miRNAs is interconnected through a network of regulatory loops driven by the FMS-like tyrosine kinase 3-internal tandem duplication (FLT3-ITD). In fact, FLT3-ITD induces the expression of miR-155 through a noncanonical mechanism of miRNA biogenesis that implicates cytoplasmic Drosha ribonuclease III (DROSHA). In turn, miR-155 down-regulates SH2-containing inositol phosphatase 1 (SHIP1), thereby increasing phosphor-protein kinase B (AKT) that in turn serine-phosphorylates, stabilizes, and activates Sprouty related EVH1 domain containing 1 (SPRED1). Activated SPRED1 inhibits the RAN/XPO5 complex and blocks the nucleus-to-cytoplasm transport of pre-miR-126, which cannot then complete the last steps of biogenesis. The net result is aberrantly low levels of mature miR-126 that allow quiescent leukemia blasts to be recruited into the cell cycle and proliferate. Thus, miR-126 down-regulation in proliferating AML blasts is downstream of FLT3-ITD­dependent miR-155 expression that initiates a complex circuit of concatenated regulatory feedback (i.e., miR-126/SPRED1, miR-155/human dead-box protein 3 [DDX3X]) and feed-forward (i.e., miR-155/SHIP1/AKT/miR-126) regulatory loops that eventually converge into an output signal for leukemic growth.

Sequence coevolution and structure stabilization modulate olfactory receptor expression.

Olfactory receptors (ORs) belong to class A G-protein coupled receptors (GPCRs) and are activated by a variety of odorants. To date, there is no three-dimensional structure of an OR available. One of the major bottlenecks in obtaining purified protein for structural studies of ORs is their poor expression in heterologous cells. To design mutants that enhance expression and thereby enable protein purification, we first identified computable physical properties that recapitulate OR and class A GPCR expression and further conducted an iterative computational prediction-experimental test cycle and generated human OR mutants that express as high as biogenic amine receptors for which structures have been solved. In the process of developing the computational method to recapitulate the expression of ORs in membranes, we identified properties, such as amino acid sequence coevolution, and the strength of the interactions between intracellular loop 1 (ICL1) and the helix 8 region of ORs, to enhance their heterologous expression. We identified mutations that are directly located in these regions as well as other mutations not located in these regions but allosterically strengthen the ICL1-helix 8 enhance expression. These mutants also showed functional responses to known odorants. This method to enhance heterologous expression of mammalian ORs will facilitate high-throughput "deorphanization" of ORs, and enable OR purification for biochemical and structural studies to understand odorant-OR interactions.

Roadmap on plasticity and epigenetics in cancer.

The role of plasticity and epigenetics in shaping cancer evolution and response to therapy has taken center stage with recent technological advances including single cell sequencing. This roadmap article is focused on state-of-the-art mathematical and experimental approaches to interrogate plasticity in cancer, and addresses the following themes and questions: is there a formal overarching framework that encompasses both non-genetic plasticity and mutation-driven somatic evolution? How do we measure and model the role of the microenvironment in influencing/controlling non-genetic plasticity? How can we experimentally study non-genetic plasticity? Which mathematical techniques are required or best suited? What are the clinical and practical applications and implications of these concepts?

Synthetic data generation with probabilistic Bayesian Networks.

Bayesian Network (BN) modeling is a prominent and increasingly popular computational systems biology method. It aims to construct network graphs from the large heterogeneous biological datasets that reflect the underlying biological relationships. Currently, a variety of strategies exist for evaluating BN methodology performance, ranging from utilizing artificial benchmark datasets and models, to specialized biological benchmark datasets, to simulation studies that generate synthetic data from predefined network models. The last is arguably the most comprehensive approach; however, existing implementations often rely on explicit and implicit assumptions that may be unrealistic in a typical biological data analysis scenario, or are poorly equipped for automated arbitrary model generation. In this study, we develop a purely probabilistic simulation framework that addresses the demands of statistically sound simulations studies in an unbiased fashion. Additionally, we expand on our current understanding of the theoretical notions of causality and dependence / conditional independence in BNs and the Markov Blankets within.

Syntaxin 4 Enrichment in ß-Cells Prevents Conversion to Autoimmune Diabetes in Non-Obese Diabetic (NOD) Mice.

Syntaxin 4 (STX4), a plasma membrane-localized SNARE protein, regulates human islet ß-cell insulin secretion and preservation of ß-cell mass. We found that human type 1 diabetes (T1D) and NOD mouse islets show reduced ß-cell STX4 expression, consistent with decreased STX4 expression, as a potential driver of T1D phenotypes. To test this hypothesis, we generated inducible ß-cell-specific STX4-expressing NOD mice (NOD-ißSTX4). Of NOD-ißSTX4 mice, 73% had sustained normoglycemia vs. <20% of control NOD (NOD-Ctrl) mice by 25 weeks of age. At 12 weeks of age, before diabetes conversion, NOD-ißSTX4 mice demonstrated superior whole-body glucose tolerance and ß-cell glucose responsiveness than NOD-Ctrl mice. Higher ß-cell mass and reduced ß-cell apoptosis were also detected in NOD-ißSTX4 pancreata compared with pancreata of NOD-Ctrl mice. Single-cell RNA sequencing revealed that islets from NOD-ißSTX4 had markedly reduced interferon-γ signaling and tumor necrosis factor-α signaling via nuclear factor-κB in islet ß-cells, including reduced expression of the chemokine CCL5; CD4+ regulatory T cells were also enriched in NOD-ißSTX4 islets. These results provide a deeper mechanistic understanding of STX4 function in ß-cell protection and warrant further investigation of STX4 enrichment as a strategy to reverse or prevent T1D in humans or protect ß-cell grafts.

RAMP2-AS1 Regulates Endothelial Homeostasis and Aging.

The homeostasis of vascular endothelium is crucial for cardiovascular health and endothelial cell (EC) aging and dysfunction could negatively impact vascular function. Leveraging transcriptome profiles from ECs subjected to various stimuli, including time-series data obtained from ECs under physiological pulsatile flow vs. pathophysiological oscillatory flow, we performed principal component analysis (PCA) to identify key genes contributing to divergent transcriptional states of ECs. Through bioinformatics analysis, we identified that a long non-coding RNA (lncRNA) RAMP2-AS1 encoded on the antisense of RAMP2, a determinant of endothelial homeostasis and vascular integrity, is a novel regulator essential for EC homeostasis and function. Knockdown of RAMP2-AS1 suppressed RAMP2 expression and caused EC functional changes promoting aging, including impaired angiogenesis and increased senescence. Our study demonstrates an integrative approach to quantifying EC aging based on transcriptome changes, which also identified a number of novel regulators, including protein-coding genes and many lncRNAs involved EC functional modulation, exemplified by RAMP2-AS1.